Number 9: Salomon Smith Barney
Salomon issued a report this morning on Micron: --SUMMARY:--Micron Technology--Semiconductors *Micron reported fiscal Q1 (November) quarter EPS of $1.19 (versus a loss of ($0.19)) on revenues of $1.6 billion (up 100% year over year and up 47% quarter over quarter). EPS was in-line with the "whisper number", though well ahead of our estimate of $0.65.*Unit shipments were flat in the qtr., lower than we anticipated, while 64Mb prices at $9.50, were in-line. Costs were much lower, contributing to 51% gross margins, well ahead of our 40% estimate.*Unit inventories rose sharply in the quarter, which probably presages further price declines ahead. We are trimming our Q2 (Feb) ASP assumption from $9.00 to $8.50, with our fiscal 2000 estimate reduced by about $0.20 to $8.23; 2001 is unchanged at $6.43.*Given the $0.55 upside, we are raising our 2000 EPS estimate, but only by about $0.20 to $3.70; we are trimming fiscal 2001 from $4.75 to $4.15, based on lower margins.
--OPINION:------------------------------------------------------------------ We expect DRAM prices to continue to decline into the first half, at least on a seasonal basis. Under this environment, we retain our 3S (Hold) rating, with a target price of $70. Big report on higher prices, gross margins. Prices in the quarter nearly doubled from the previous quarter to about $9.75, up from $5.10 the previous quarter, but in line with our $9.50 estimate. Earnings were well ahead of our expectations, however, largely because we did not fully figure in the huge operating leverage generated through the price increases. Gross profits grew over 250% quarter over quarter, with gross margins rising over 30-percentage points from 21% to 51%, which was still well above our forecast of 40%. The importance of pricing is underscored by the fact that virtually all profit increase in the quarter came from higher prices. Megabit shipments flat, while unit inventories grew. We were surprised that Micron did not increase their unit shipments in the quarter, though manufactured units (megabits) grew by about 35%. The increase in unit production went to replenish inventories, which were worked down to virtually nothing in fiscal Q4. If inventories fell to two weeks of 64Mb equivalent sales in Q4, and production increased by 35% while shipments out were flat, the increase in unit inventory could have been as much as 40-50 million units, or 4-5 weeks of unit sales. An estimate of 6-7 weeks of inventory for Q2 is probably too high, but 5 weeks of inventory would not seem a bad estimate. Five weeks is not excessive, but it will signal to brokers and PC OEMs that DRAM makers are once again building inventories, and they will certainly press hard for further contract price concessions. This foots with most DRAM manufacturer's forecasts that DRAM pricing will continue to slide into Q1. Demand strong in the quarter, but Q1 outlook for seasonal norm. Given the continuing rapid increase in unit shipments due to aggressive shrink strategies, the only factor that could firm up prices would be much stronger than anticipated demand in Q1. Officially, the company is not committing itself to more than a seasonally in-line quarter, which would imply a 10% sequential drop in unit shipments. However, off-line, management is hearing of continuing PC component shortages, which could suggest a stronger than seasonal Q1 as U.S. corporate lock-downs come off. We too, are looking for a better than seasonal Q1, but do not believe it will be enough to prevent further DRAM price declines. A little more confusing with regard to commentary on the call was the statement that Micron maintained market share in the last quarter, though did not increase unit shipments out much. The explanation was that Back-to-School resulted in a huge spike in demand, which may have been tempered in fiscal Q1 by the Taiwanese earthquake, price elasticity from higher prices, and a slowdown in U.S. corporate demand. Rodney Dangerfield visits Boise. The frustrating part for any investor that wants to bet on Micron as the DRAM leader benefiting from a cyclical upturn is the fact that the upturn has not really materialised. At least not the way it did last cycle. From October 1990 to October 1995, Micro n's stock rose more than 3,000%. That may not seem like much these days, but it was a huge return for those who believed in outstanding management in Boise, outstanding products, and a reluctance by Asian suppliers to invest in greenfield fabs. Sound familiar? The difference between then and now is the fact that the whole semiconductor world has literally adopted techniques first developed in Boise: essentially an aggressive shrink strategy that allows for greater production with less cost. This shrink strategy has radically changed the DRAM industry wherein lower levels of capital spending have been sufficient to meet increased market demand for product. Nevertheless, we are betting in the long-term on the Boys from Boise. If we can ever get stable pricing. |
